Artificial intelligence and machine learning have the potential to revolutionise healthcare in general, but notably in the area of precision medicine.
Precision medicine is frequently distinguished from the commonly used term “personalised medicine,” which merely refers to adapting therapies to the individual patient. Precision medicine, on the other hand, uses machine learning to analyse the genetic material of patients with rare diseases. Artificial intelligence (AI) searches for patterns in data to uncover common traits, which pharmaceutical companies then utilise to design medications that are tailored to the individual need.
Palo Alto, California-based Endpoint Health is one player in this space looking to tap the potential machine learning has for precision medicine. Today, it announced it has received $52 million series A equity and debt financing – a big jump from its first funding round of $12 million in 2020.
The fresh funds will be used to develop the company’s Precision-First platform and add initiatives for chronic immune-mediated disorders to its therapeutic pipeline. The funds will also help Endpoint Health investigate the use of Antithrombin III, a human plasma-derived medicine, in patients with sepsis, a life-threatening medical emergency that arises when the body’s immune response to infection becomes dysregulated.
Clinical studies can be completed faster and at a cheaper cost thanks to artificial intelligence.
A Phase II clinical trial is intended to investigate the use of Antithrombin III in patients with sepsis. Endpoint Health used ML-driven insights to develop a blood test that will identify patients who will be most helped by Antithrombin III. To develop the test, which identifies patients with a particular form of sepsis, the Precision First platform analyzed RNA from sepsis patients, using machine learning to “look for underlying patterns that are unique and different that a normal human couldn’t see,” said Jason Springs, Endpoint Health chief executive officer.
The immune reactions of patients with sepsis differ, so the platform discovers subgroups and identifies what people in those groups have in common.
“The immune system is very complex, so the machine learning platform lets the computer pull apart that complexity in a way no human could,” Springs said.
“Nobody could look at 10,000 data points for 1,000 patients and quickly understand if there are one or two different clusters of patients that look like each other – one person can’t process that much information at once,” he said. “But the computer will keep crunching and it can handle the amount of data required to understand these illnesses.”
This initial analysis determined whether to develop a blood test for a group of sepsis patients “that the computer told us were unique,” Springs said, adding that the test identifies patients within a sepsis subtype. Those patients can take part in the upcoming Phase II clinical trials for the sepsis drug. Identifying patients in this way makes clinical trials shorter and less expensive, as researchers can choose the participants they think most likely to respond to their therapy.
The use of artificial intelligence to analyze genetic material in order to match patients to medications
The AI analysis of genetic samples from sepsis patients who agree to contribute blood samples kicks off the entire procedure.
“If we receive 1,000 samples, we’ll know in a month or two if we notice any patterns,” Springs added.
In the future, the test will be able to identify sepsis subgroups, allowing doctors to match patients with the appropriate treatment for their condition. Springs emphasised that such a match can’t occur fast enough. Small blood clots are dispersed throughout the body in some sepsis patients, causing disseminated intravascular coagulation (DIC). The situation is quite perilous.
“Our hope is that our therapy will have a good chance of resolving that problem,” he said.
Other companies are also using AI tools to deliver precision medicine. According to The Journal of Precision Medicine, embracing digitization is the “the key to enable and operationalize both standardization and personalization in health care.” Tools can help affordably create and deliver complex, patient-specific pharmaceutical or medical devices.
Other precision medicine companies using AI include Synapse, GNS Healthcare andTempus, which in 2020 announced an additional $100 million financing, bringing its financing total to $600 million since its 2015 inception. Tempus focuses on cancer treatment, though it devoted resources to COVID-19.
Endpoint Health’s own recent funding round is impressively large because the company’s technology attracts investors.
“Investors are interested in technology that can analyze and understand large sets of unique patient data and gain insights that could speed clinical development of therapies that are much more likely to succeed in patients,” Springs says.
The AI healthcare market is expected to exceed $35 billion by 2025.
According to a September 2021 analysis from EIT Health and McKinsey, the United States dominates the list of firms with the largest VC funding in healthcare AI to date, as well as having the most completed AI-related healthcare research projects and trials.
The global AI healthcare market spend is anticipated to reach over $35 billion by 2030, growing by 24% from $2 million in 2019, according to the BIS Research market intelligence firm.
“We are in the very early days of our understanding of AI and its full potential in healthcare, in particular with regards to the impact of AI on personalization,” according to the report, which predicts precision medicine will grow to offer medicine tailored to every patient’s unique need.
While that is in the future, Springs can attest to the fact that precision medicine is already here. And the potential for AI tools in precision medicine will continue to skyrocket.